-
-
Notifications
You must be signed in to change notification settings - Fork 2
/
draft.qmd
151 lines (113 loc) · 12.9 KB
/
draft.qmd
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
# The Fantasy Draft {#sec-draft}
## Getting Started {#sec-draftGettingStarted}
### Load Packages {#sec-draftLoadPackages}
```{r}
```
## Types of Fantasy Drafts {#sec-draftTypes}
There are several types of drafts in fantasy football.
The most common types of drafts are snake drafts and auction drafts.
### Snake Draft {#sec-draftSnake}
In a snake draft, the participants (i.e., managers) are assigned a draft order.
In the first round, the managers draft in that order.
In the second round, the managers draft in reverse order.
It continues to "snake" in this way, round after round, so that the person who has the first pick in a given round has the last pick in the next round, and whoever has the last pick in a given round has the first pick in the next round.
### Auction Draft {#sec-draftAuction}
In an auction draft, the managers are assigned a nomination order and there is a salary cap (e.g., $200).
The first manager chooses which player to nominate.
Then, the managers bid on that player like in an auction.
In order to bid, the manager must raise the price by at least $1.
If two managers want to obtain the same player, they may continue to raise the amount until one manager backs out and is no longer to bid by raising the price.
The highest bidder wins (i.e., drafts) that player.
Then, the second manager nominates a player, and the managers bid on that player.
This process repeats until all teams have drafted their allotment of players.
### Comparison {#sec-draftComparison}
Snake drafts are more common than auction drafts.
Snake drafts tend to be quicker than auction drafts.
However, auction drafts are more fair than snake drafts.
In an auction draft, unlike a snake draft, all players are available to all teams.
For instance, in a snake draft, the first 9 players drafted are unavailable to the 10th pick of the first round.
So, if you have the 10th pick and want the top-ranked player, this player would not be available to you in the snake draft.
However, in the auction draft, every player is available to every manager, so long as the manager is able and willing to bid enough.
## Draft Strategy {#sec-draftStrategy}
### Overview {#sec-draftStrategyOverview}
There is no one "right" draft strategy.
As noted by @Lee2022 in their analysis of fantasy drafts, the effectiveness of any draft strategy depends on the strategies of the other managers in the league.
Sometimes it works best to "zig" when everyone else is "zagging".
For instance, @Lee2022 found that the most successful draft strategies in terms of roster composition (i.e., the number of players for each position) were not the most common draft strategies.
For instance, if you notice that everyone else is drafting Wide Receivers, this may mean that other managers are over-valuing Wide Receivers, and this could be a nice opportunity to draft a Running Back for good value.
In general, you will first want to generate the rankings you will use to select which players to prioritize.
You may generate your rankings based one or more of the following:
- your [evaluation of players](player-evaluation.qmd)
- [expert or aggregated rankings](#sec-fantasyValueRankings)
- [layperson rankings](#sec-fantasyValueADPAAV)
- players' [Average Draft Position](#sec-fantasyValueADPAAV) (ADP) in other league drafts (for [snake drafts](#sec-draftStrategySnake))
- players' [Average Auction Value](#sec-fantasyValueADPAAV) (AAV) in other league drafts (for [auction drafts](sec-draftStrategyAuction))
- [expert or aggregated projections](#sec-fantasyValueProjections)
- [indices derived from rankings and projections](#sec-fantasyValueIndices)
As described in @sec-wisdomOfCrowd, there can be benefits of leveraging the [wisdom of the crowd](#sec-wisdomOfCrowd) by using rankings or projections that are averaged across many people and perspectives.
@sec-fantasyValueSources describes where to obtain [aggregated rankings](#sec-fantasyValueRankings), [aggregated projections](#sec-fantasyValueProjections), [ADP](#sec-fantasyValueADPAAV), and [AAV](#sec-fantasyValueADPAAV) data.
As described in @sec-judgmentVsActuarial, there are also benefits to using the [actuarial approach](#sec-judgmentVsActuarial) to prediction rather than (merely) using judgment.
It is not sufficient to compare players in terms of projected points because some positions have more depth than other positions.
Some positions show *positional scarcity*—that is, a limited number of high performing players.
An important concept in the draft is "[dropoff](#sec-fantasyValueDropoff)", which is described in @sec-fantasyValueDropoff.
[*Dropoff*](#sec-fantasyValueDropoff) at a given position, is the difference—in terms of projected fantasy points—between (a) the best available player remaining at that position and (b) the second-best available player remaining at that position.
If there is a bigger [dropoff](#sec-fantasyValueDropoff) at a given position, there may be greater value in drafting the top player from that position.
For instance, consider the following scenario:
"Quarterback A" is projected to score 325 points, and "Quarterback B" is projected to score 320 points.
"Tight End A" is projected to score 230 points, and "Tight End B" is projected to score 150 points.
In this example, there is a much greater [dropoff](#sec-fantasyValueDropoff) for Tight Ends than there is for Quarterbacks.
Thus, even though "Quarterback A" is projected to score more points than "Tight End A", "Tight End A" may be more valuable, relatively, because there is still a good Quarterback available if someone else drafts "Quarterback A".
In general, Kickers and Defenses tend to have the lowest dropoff (i.e., the lowest expected drop in fantasy points) by positional rank, so it makes sense to draft Kickers and Defenses late in the draft [@Lee2022].
Defenses, in particular, appear to be among the least predictable of the positions [@Lee2022].
Another important concept is a player's [value over a typical replacement player](#sec-fantasyValueVORP) at that position (shortened to "value over replacement player"; VORP), which is described in @sec-fantasyValueVORP.
A player's [value over a typical replacement player](#sec-fantasyValueVORP) provides a way to more fairly compare (and thus rank) players across different positions.
Another important concept is a player's [uncertainty](#sec-fantasyValueUncertainty), which is described in @sec-fantasyValueUncertainty.
In both [snake](#sec-draftStrategySnake) and [auction](#sec-draftStrategyAuction) draft formats, your goal is to draft the team whose weekly starting lineup scores the most points and thus the collection of players with the greatest [VORP](#sec-fantasyValueVORP).
For your starting lineup, it may make sense—especially with your earliest selections—when comparing two players with equivalent [VORP](#sec-fantasyValueVORP), to prioritize players with higher [consistency](#sec-evalHistoricalConsistency) and lower [uncertainty](#sec-fantasyValueUncertainty), because they may be considered "safer" with a higher floor.
However, when drafting players for your bench, it make make more sense to prioritize high-risk, high reward players with greater [uncertainty](#sec-fantasyValueUncertainty), because they may have a higher ceiling.
Players with a higher ceiling have a potential to be "sleepers"—players who are valued low (i.e., with a high [ADP](#sec-fantasyValueADPAAV) or low [AAV](#sec-fantasyValueADPAAV)) and who outperform their valuation.
Note that, although players with greater [uncertainty](#sec-fantasyValueUncertainty) are high-risk, high-reward players, selecting this kind of a player for your bench (i.e., in a late round or for a small cost) is a *lower* risk selection, because you have less to lose with later/lower-cost picks.
That is, even though the *player* is higher risk, selecting a higher risk player for your bench is a lower risk *decision*.
The Spurs in the National Basketball Association (NBA) were well-reputed for [excelling in this draft strategy](https://harvardsportsanalysis.org/2013/11/beating-the-nba-draft-does-any-team-outperform-expectations/) (archived at <https://perma.cc/X7NW-WZC6>).
They frequently used their second-round picks to draft high-risk, high-reward players.
Sometimes, the secound round pick was a bust, but they have little to lose with a failed second round pick.
Other times, their second round picks—including Willie Anderson, DeJuan Blair, Goran Dragic, Luis Scola, and Manu Ginóbili—greatly outperformed expectations.
Thanks, in part, to this draft strategy, the team showed strong extended success for nearly three decades from 1989 through the late-2010s.
However, the draft strategies to achieve the "optimal lineup" differ between [snake](#sec-draftStrategySnake) versus [auction](#sec-draftStrategyAuction) drafts.
One factor that is *not* included above is whether a player is on your favorite team.
Managers commonly like to draft players on their favorite teams (e.g., Cowboys, Eagles).
That is fine—fantasy football is a game.
Do what is fun for you.
However, if your goal is to select the best players, leave your allegiances at the door.
Selecting players based on their playing for your favorite team—rather than based on performance—is a form of [cognitive bias](#sec-cognitiveBias).
In general, the most consequential decisions tend to be those made early in the draft regarding the top projected players [@Lee2022].
That is, the earlier selections tend to have the greatest impact on the success of one's fantasy season.
So, make sure to spend time in your draft preparation to identify the players you want to select early in the draft.
Moreover, some teams like to hedge their bet on their top Running Backs in the case that the player were to get injured, by drafting the player's backup, a strategy known as *handcuffing*.
However, there is not strong evidence that handcuffing leads to better outcomes [@Lee2022].
In general, @Lee2022 found that teams that drafted more Running Backs and Wide Receivers tended to outperform other teams.
### Snake Draft {#sec-draftStrategySnake}
In general, your goal is to draft the team whose weekly starting lineup has the greatest [VORP](#sec-fantasyValueVORP).
Consequently, you are often looking to pick the player with the highest [VORP](#sec-fantasyValueVORP) at a given selection, while keeping in mind (a) the [dropoff](#sec-fantasyValueDropoff) of players at other positions and (b) which players may be available at subsequent picks so that you do not sacrifice too much later value with a given selection.
For instance, if a particular Quarterback has a slightly higher [VORP](#sec-fantasyValueVORP) than a particular Running Back, but the Quarterback is likely to be available at the manager's next pick but the Running Back is likely to be unavailable at their next pick, it might make more sense to draft the Running Back.
Herding behavior is common in snake drafts.
As described in @sec-wisdomOfCrowd, herding occurs when people align their behavior with others.
For instance, @Lee2022 found evidence that when Quarterbacks, Kickers, and Defenses were drafted by one team, subsequent teams were more likely to draft a player of that position.
The same was not the case for Running Backs, Wide Receivers, and Tight Ends.
### Auction Draft {#sec-draftStrategyAuction}
According to [an analysis](https://harvardsportsanalysis.wordpress.com/wp-content/uploads/2012/04/fantasyfootballdraftanalysis1.pdf) by the Harvard Sports Analysis Collective (archived at <https://perma.cc/P7RX-92UU>), the majority of the manager's salary cap should be spent on the starting lineup, and you should spend less on bench players.
This is known as the "stars and scrubs" draft strategy.
Based on the analysis, the author recommended applying a 10% premium to the top players and a 10% discount to the lower-tiered players.
The idea behind the approach is that a player on your bench does not contribute to the team's points and, thus, most players drafted to your bench do not contribute much to the team's points throughout the season.
That said, bench players can be important in the case of a starter's injury or under-performance.
So, it is recommended to draft starters with lower [uncertainty](#sec-fantasyValueUncertainty) who are safer.
In contrast to your starting lineup, you may look to draft players on your bench who have greater [uncertainty](#sec-fantasyValueUncertainty) for their high reward potential in a low-risk selection given the lower price.
An alternative to the "stars and scrubs" approach is to wait to draft more "high-value" players after other managers have over-paid for players.
In any case, having some small amount of cap left over toward the end of the draft can help you draft good value players to fill out your bench spots for cheap (e.g., $2).
Having some depth can help you offset the risk of injuries and the possibility that some of your starters may underperform their expectation, which is quite likely given how [challenging it is to predict fantasy performance](#sec-accuracyOfCrowd).
::: {.content-visible when-format="html"}
## Session Info {#sec-draftSessionInfo}
```{r}
sessionInfo()
```
:::